HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Swanson, D. J. Articles by Lewis, E. J. Search for Related Content PubMed PubMed Citation Articles by Swanson, D. J. Articles by Lewis, E. J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 4, 2911-2923, January 28, 2000

The Homeodomain Protein Arix Promotes Protein Kinase A-dependent Activation of the Dopamine -Hydroxylase Promoter through Multiple Elements and Interaction with the Coactivator cAMP-response Element-binding Protein-binding Protein^*

Douglas J. Swanson, Megumi Adachi, and Elaine J. Lewis

From the Department of Biochemistry and Molecular Biology, Oregon Health Sciences University, Portland, Oregon 97201

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The differentiation and maintenance of a neurotransmitter phenotype is guided by the interaction of exogenous cues with intrinsic genetic machinery. For the noradrenergic phenotype, these influences combine to activate the expression of the catecholaminergic biosynthetic enzymes tyrosine hydroxylase and dopamine -hydroxylase (DBH). In this study, we evaluate the molecular mechanisms by which the transcription factor Arix/Phox2a contributes to DBH gene transcription. We have evaluated the contribution of individual homeodomain binding sites in the rat DBH promoter region and find that all are essential for both basal and cAMP-dependent protein kinase A (PKA)-stimulated transcription. Using mammalian one-hybrid and two-hybrid systems, we demonstrate that recruitment of Arix to the positions of homeodomain core recognition sites 1 and 2 at 153 to 166 of the DBH gene restores complete responsiveness of the promoter to PKA in SHSY-5Y neuroblastoma and HepG2 hepatoma cells. Intracellular Arix-Arix interactions are evident and may contribute to the interdependence of homeodomain binding sites. Analysis of functional domains of Arix reveals an N-terminal activation domain and a C-terminal repression domain. The N terminus of Arix contains an amino acid motif similar to a region in Brachyury and Pax9 transcription factors. The N-terminal activation domain of Arix interacts with the transcriptional co-activator, cAMP-response element-binding protein-binding protein, which potentiates transcription from the DBH promoter in a PKA-dependent manner. The present study supports the hypothesis that the paired-like homeodomain protein, Arix, acts as a critical phenotype-specific regulator of the DBH promoter by serving as an integrator of signal-dependent transcription activators within the network of the general transcription machinery.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

One of the key goals in developmental neurobiology is to understand the mechanisms underlying the generation of neuronal diversity. At the foundation of this issue are the functional mechanisms that mediate phenotype-specific regulation of genes. Environmental cues along the path of migrating neural crest cells are thought to induce the expression of proneural transcription factors that govern genetic pathways responsible for the coordinate regulation of pan-neuronal and phenotype-specific markers (1-3). Members of the Arix/Phox2 family of paired-like homeodomain transcription factors are thought to be critical components of the genetic pathways involved in the differentiation of noradrenergic characteristics in neurons (4-7). The noradrenergic phenotype is typified by the coexpression of the enzymes critical for noradrenaline synthesis, tyrosine hydroxylase, and dopamine- hydroxylase (DBH).¹ Mice with null mutations of Phox2a or Phox2b exhibit the conspicuous failure of central or peripheral noradrenergic populations to differentiate (4). In cultured neural crest cells, bone morphogenetic protein 2 was shown to induce Arix/Phox2a and to eventually express markers of panneuronal differentiation (5), but co-stimulation of the cAMP/PKA system is required to induce the coexpression of tyrosine hydroxylase and DBH expression at detectable levels (7). Thus, it seems that phases of the genetic programs regulating neurotransmitter phenotype may require additional instructional cues to push differentiation toward a specific fate and drive sustained expression phenotype-specific genes.

A deeper understanding of the role genetic pathways in regulating cellular phenotype is limited by our knowledge of how these transcription networks function to directly regulate phenotype-specific target genes. Although genetic manipulation studies strongly implicate the Arix/Phox2 family in the regulation of the noradrenergic phenotype, they have not demonstrated whether these proteins are directly responsible for the transcriptional activation of the DBH or tyrosine hydroxylase genes. Cell transfection of DNAs into cultured cells and biochemical analyses of DNA-protein interactions have shown that Arix/Phox2 family members regulate the noradrenergic phenotype by direct activation of the promoters of both tyrosine hydroxylase and DBH genes (8-12).² Furthermore, activation of the DBH promoter by Arix is synergistically stimulated by the cAMP/PKA pathway (9). This result is consistent with findings of Lo et al. (7), where bone morphogenetic protein 2 induction of DBH gene expression in neural crest cells requires not only the function of Arix/Phox2 proteins but also stimulation of the PKA pathway. Thus, it appears that Arix functions as a node for integration of multiple cues within the genetic pathway regulating expression of noradrenergic specific genes.

It is evident that Arix/Phox2a is a necessary factor for the selective expression of the DBH gene in noradrenergic neurons. However, little is known about the mechanism(s) that underlie Arix/Phox2a-mediated transcription from the DBH promoter. The goal of this study was to elucidate the functional action of Arix at the DBH promoter as it regulates the DBH gene. We have defined an enhancer element, the DB1 enhancer, within the rat DBH promoter that is essential for tissue-specific regulation, Arix-dependent activation, and PKA-induced modulation (9, 13). Within this enhancer is found a CRE/AP1-like element located adjacent to two homeodomain core recognition (HD) sites, displaying a characteristic central ATTA motif. The two HD sites of the DB1 enhancer (HD1/2) and a third, more proximally located HD site (HD3) can serve as binding sites for members of the Arix/Phox2 family (8-12), and these Arix/Phox2 proteins bind to these sites in nuclear extracts from catecholaminergic cells. The CRE/AP1 site binds PKA-induced AP1 transcription factors, c-Fos and c-Jun, and is also necessary for the synergistic promoter activation by PKA and Arix (9, 10). Taken together, these findings suggest that Arix and PKA-induced AP1 proteins synergistically drive DBH transcription by directly activating the DBH promoter in a combinatorial fashion. The precise role of Arix and how it functions through the HD sites in the promoter is still uncertain.

An emerging theme in transcriptional activation is the assembly of multiprotein regulatory complexes at enhancers and promoters of target genes (14). These complexes are formed through multiple protein-DNA and protein-protein interactions. The mechanisms that underlie transcriptional synergy through multiple transcription factors may then rely on either the cooperative binding of these factors to the DNA or the stable co-recruitment of coactivators or components of the general transcription machinery (reviewed by Tijan and Maniatis (14)). The transcriptional coactivators, CBP/p300, are large, multifunctional proteins, containing histone acetyltransferase activity (reviewed by Goldman et al. (15)). These transcriptional activators are believed to facilitate transcription by remodeling chromatin structure (16, 17) and by recruiting components of the general transcription machinery to the promoter (18, 19). CBP/p300 contain several domains necessary for interaction with transcriptional activators including CREB (20), c-Fos (21), and c-Jun (22) as well as the general transcription factors, TFIIB and TFIID (18, 19). The recruitment of CBP/p300 proteins to the promoter may then be stabilized through multiple contacts with transcriptional activators distributed at different sites along the promoter. A stable coactivator complex could then greatly facilitate transcription and would be manifest as a much stronger synergistic response.

To understand the mechanism by which Arix regulates transcription of the DBH and tyrosine hydroxylase promoters, it is necessary to define the transcriptional activation domains and interacting coactivator proteins that participate in Arix function. We have shown that Arix possesses an activation domain that functions in a DBH promoter context-dependent manner. Intracellular Arix-Arix interactions are evident and may influence binding to the three homeodomain recognition sites in the DBH promoter-proximal region. It appears that the functional synergism of Arix with PKA involves the coactivator, CBP, which interacts with the Arix activation domain. These results demonstrate that Arix is a tissue-specific transcriptional activator, which regulates DBH promoter responsiveness by combining with other factors to recruit coactivators and the general transcription machinery directly to the DBH promoter, thereby facilitating basal and activated transcription of the DBH gene.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- Human HepG2 hepatoma cells were cultured in minimum Eagle's medium supplemented with 10% fetal bovine serum (Hyclone), 1% nonessential amino acids, and 110 mg/liter sodium pyruvate. Human SHSY-5Y neuroblastoma cells were cultured in a 1:1 ratio of F-12 nutrient mixture and minimum Eagle's medium including 10% fetal bovine serum. All cell lines were maintained at 37 °C in an atmosphere of humidified air containing 5% CO₂.

Reporter Plasmid Constructs-- The construction of the reporter plasmid containing the promoter and 5'-flanking sequence of DBH gene cloned adjacent to the bacterial chloramphenicol acetyltransferase (CAT) transcription unit (DBH(232)-CAT) was described previously (13). The DBH(232)-Luc reporter plasmid was constructed by insertion of a HindIII-MluI fragment from DBH(232)-CAT into the luciferase reporter vector, pGL3-Basic (Promega). This reporter construct contains the proximal DBH promoter (232/+12) including the DB1 enhancer region, TATA-like sequence, and the transcription start site (23), linked upstream of the coding sequence of firefly (Photinus pyralis) luciferase (FL). Mutations of the DBH(232) reporter were made in the DBH(232)Luc reporter by oligonucleotide-directed mutagenesis (Quick Change, Promega). The HD3m reporter construct was made using the complementary oligonucleotide pair CACCAGACAAATGTctagAGGTACAGCC (base pair substitutions in lowercase type) derived from the proximal HD3 and flanking sequences (80 to 52; see Fig. 1A) in which the core ATTA site was disrupted. The 2HDm reporter construct was made using the complementary oligonucleotide pair ATGTCCATGCGTCATacGTGTCAccTAGGG (base pair substitutions in lowercase) derived from the DB1 enhancer sequence (180 to 151; see Fig. 1A) in which both homeodomain core ATTA sites were disrupted. The 2HDgal reporter construct was made using the complementary oligonucleotide pair CCATGCGTCATTcggagtactgtcctccgGATCGGAGC (derived from bp 176 to 139) containing the pMH100 Gal4-UAS (in lowercase) (24), which acts as a high affinity binding site for the Gal4-DNA binding domain (Gal-DBD). This mutation disrupts the sequence containing the HD1 and HD2 sites within the DB1 enhancer leaving the CRE/AP1 site intact. The 3HDm and 3HDgal reporters were made by insertion of an EcoRI/BglII fragment containing the HD3m mutation into similarly digested 2HDm and 2HDgal constructs respectively, resulting in constructs containing all three HD core mutations. All reporter constructs were subjected to automated sequencing to confirm the accuracy of polymerase chain reaction amplification throughout the wild-type and mutated promoter sequences.

To construct 5Gal-DBH(125)-Luc and 5Gal-DBH(62)-Luc constructs, a fragment containing five tandem copies of the Gal4-UAS site was transferred from the plasmid vector 5×Gal4-TATA-Luciferase (25) to the DBH reporter vectors. For the 5Gal-DBH(125)-Luc and 5Gal-DBH(62)-Luc constructs, an XbaI site was engineered in the DBH proximal promoter, and the Gal4-UAS sites from 5×Gal4-TATA-luciferase were inserted into KpnI/XbaI-digested reporter vectors at 125 and 62 bases upstream of the transcription start site, respectively. Thus, these reporter constructs contain the 5Gal-UAS sites and the DBH proximal promoter sequence including the DBH-TATA box and other respective upstream elements necessary to drive the luciferase-reporter transcription unit (illustrated in Fig. 2A). The functions of these constructs were verified by activation by a Gal-DBD:VP16-AD fusion construct (a gift from Dr. Richard Maurer), which recruits the strong VP16 activation domain to the promoter through the Gal4-UAS sites.

Expression Plasmid Constructs-- The RSV-PKA expression construct was a generous gift from Dr. Richard Maurer (Oregon Health Sciences University; previously described in Ref. 26). The pCBP-HA expression plasmid was a generous gift from Dr. Richard Goodman (Oregon Health Sciences University; previously described in Ref. 27). The construction of RSV-Arix has been previously described (9).

The cDNA for human NBPhox was constructed by a combination of polymerase chain reaction and genomic cloning techniques. DNA encoding the N-terminal and homeodomain portion of NBPhox were cloned by reverse transcriptase polymerase chain reaction of RNA from SHSY-5Y neuroblastoma cells, using primers 5'-TGCTCTAGAGACCTCAGACAAGG-3' and 5'-GAAGAGTCAGACTTTTTGCCCG-3', encompassing bases 174-896 of the original NBPhox cDNA (28). The C-terminal and 3'-untranslated segments are encoded within one exon and were identified by isolation of a genomic clone of human NBPhox. The two segments were recombined by cloning techniques, generating an NBPhox cDNA corresponding to bases 177-1831 of the original cDNA. This cDNA contains the entire coding sequence, 185 bases of 5'-untranslated and 533 bases of 3'-untranslated regions, and is cloned into vector pcDNA3.

Hemagglutinin (HA)-tagged expression constructs for wild-type and truncated Arix were made using the HA6.1 expression vector developed and generously provided by Dr. Paul Shapiro. HA6.1 consists of a pcDNA3 (Invitrogen) backbone with an N-terminal HA epitope tag and MCS for in-frame fusion with target proteins. The precise construction of HA-tagged wild-type Arix (aa 1-280) and ArC (aa 1-151), ArN (aa 84-280), and ArHD (aa 84-151) has been described elsewhere.² Each construct was created with an EcoRI site 3' to the HA epitope that was used to create in-frame Gal-DBD fusions described below.

The vectors Gal-DBD and VP16-AD (gifts from Dr. Richard Maurer) were used as the parent for making Gal-DBD fusion constructs with full-length and truncated Arix expression constructs. The Gal-DBD vector consists of a backbone pcDNA3 vector (Invitrogen) using the cytomegalovirus promoter to drive an expression cassette containing the zinc finger DNA binding domain of GAL4 (aa 1-147), derived from the plasmid pSG424 (25, 29), as an N-terminal fusion with target proteins. The VP16-AD vector also consisted of a pcDNA3 backbone containing the herpes simplex VP16-activation domain (AD, aa 402-479) as an N-terminal fusion with target proteins. Full-length Arix-VP16 was created by insertion of an EcoRI-NotI fragment from HA-Arix into the VP16-AD parent vector. Full-length Gal-Arix (aa 1-281), Gal-ArC (aa 1-150), and Gal-ArHD (aa 84-150) constructs were made by insertion of EcoRI-NotI fragments from HA-Arix, HA-ArC, and HA-ArHD, respectively, into the Gal-DBD vector. Gal-NAr (aa 1-101) was created by insertion of a linker between the Bpu1102 and XbaI sites of full-length Gal-Arix, creating a stop codon after residue Lys¹⁰¹ and removing much of the 3'-untranslated region of Arix. Gal-ArC (aa 148-281) was created by insertion of a linker between the EcoRI and BssHII sites of full-length Arix, creating an in-frame fusion of the Gal-DBD with residue Gln¹⁴⁸. Gal-nAr (aa 26-281) was created by insertion of a linker between the EcoRI and NarI sites of full-length Arix, resulting in the insertion of a proline residue between the Gal-DBD and residue Ala²⁶ of Arix. All expression constructs produced in vitro translated proteins of the appropriate predicted sizes.

Transient Transfection Analyses-- DNA constructs used for transfection were purified using the Promega Wizard kit. Following purification according to the manufacturer's procedures, DNA was ethanol-precipitated in the presence of ammonium acetate and then resuspended in sterile water. SHSY-5Y and HepG2 cells were transfected by the calcium phosphate method as described previously (13, 30). Cells were plated in six-well plates at a density of 0.5-1 × 10⁶ cells/well and transfected with a total of 3-4 µg of plasmid DNA (as indicated in each figure). All transfections using the DBH-Luc based reporters included 0.5 µg of the pRL-null reporter construct (Promega) as a control for transfection efficiency. This vector is a promoterless reporter containing the transcription unit of Renilla luciferase (RL). We have found that the pRL-null reporter gives measurable levels of RL expression that are consistent across experiments, and expression is not stimulated by cotransfection with PKA.³ Thus, the use of this promoterless reporter is ideal for normalization in experiments measuring the effects of PKA stimulation. Cells were harvested 48 h after transfection and lysed in passive lysis buffer. 20 µl of cleared lysates were sequentially assayed for firefly luciferase and Renilla luciferase using the Dual-Luciferase Assay System (Promega) according to the manufacturer's instructions. Reporter activities are calculated as the ratio of light units of firefly luciferase to light units of Renilla luciferase (FL/RL).

For CBP experiments, HepG2 cells were plated in 100-mm plates at a density of 3 × 10⁶ cells/plate. Transient transfections were performed using calcium phosphate with a total of 20 µg of plasmid. Cells were harvested 48 h after transfection, and aliquots of cell extracts were assayed for protein content and CAT activity (3). CAT activity was normalized to total lysate protein, since previous experimental results indicated a stimulatory effect of cAMP and PKA on the RSV promoter of RSV-luciferase (9).

In Vitro Protein Interaction Assays-- Glutathione S-transferase (GST) fusion constructs of fragments of CBP, GST-CBP-(1-450), GST-CBP-(450-682), GST-CBP-(1091-1330), GST-CBP-(1679-1874), were a generous gift from Drs. Hua Lu (Oregon Health Sciences University) and Dick Goodman (Vollum Institute). GST-CBP fusion proteins were prepared and bound to glutathione-agarose (Sigma) following the procedure described by Ausbel et al. (32). Each preparation was analyzed by SDS-polyacrylamide gel electrophoresis to verify protein integrity and estimate the concentration of protein load on beads. Agarose beads containing approximately equal amounts of GST fusion protein were resuspended in binding buffer containing 50 mM potassium phosphate, pH 7.5, 150 mM KCl, 1 mM MgCl₂, 10% glycerol, 1% Triton X-100). Test proteins were labeled with [³⁵S]methionine by in vitro translation using the TNT T7 coupled reticulocyte lysate system (Promega) following the manufacturer's procedure. Equal amounts of labeled protein were incubated with the fusion protein-bead complexes for 1 h with rotation at 4 °C. Affinity complexes were washed five times in binding buffer, resuspended in sample buffer, boiled for 5 min, and analyzed by SDS-polyacrylamide gel electrophoresis. Radioactively labeled proteins were then visualized by autoradiography.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

HD Sites in the DBH Promoter Critical for Tissue-specific and Arix-mediated Transcription-- The rat DBH promoter region contains multiple HD sites (see Fig. 1A). Two HD sites are found within the previously identified DB1 enhancer region (13), approximately 160 bp from the transcriptional start site. This pair of HD sites, designated HD1/2, is positioned adjacent to the functionally critical CRE/AP1-like site (9). A third, more proximally located HD (HD3) is found 60 bp upstream of the start site. We have previously demonstrated that mutations within the DB1 enhancer that disrupted HD1/2 sites decreased basal and stimulated transcription from the DBH promoter (8, 9). These mutations, however, also disrupted the adjacent AP1-like site. To evaluate the distinct contribution of the HD1/2 pair and the HD3 site to cell type-specific DBH promoter regulation, small (3-4-bp) mutations were used to disrupt the ATTA core sequence of each of these HD elements. The noradrenergic neuroblastoma cell line, SHSY-5Y, and the hepatocarcinoma cell line, HepG2, served as models to compare cell type specificity of DBH transcription in transient transfection analyses. SHSY-5Y cells express the DBH gene as well as Arix and NBPhox. HepG2 cells are negative for neuronal markers, including Arix and NBPhox, providing a sufficient cellular background within which to test the effects of exogenously expressed Arix on DBH promoter regulation.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Multiple HD sites are necessary for tissue-specific regulation of the DBH promoter. A, a schematic representation of the wild-type DBH(232) and HD mutant promoters with a sequence detail of the regulatory elements, CRE/AP1 (shaded), and homeodomain core sites (white boxes), HD1, HD2, and HD3. The DB1 enhancer element (underlined) has been shown to be critical for tissue-specific and cAMP-regulated transcription. The mutated HD core sites in each promoter construct are designated by striped boxes. Transcriptional activity analyses of the wild-type and the mutant DBH(232)-Luc reporter constructs are shown in SHSY-5Y (B) and HepG2 (C) cells. SHSY-5Y cells were cultures in six-well plates and transiently transfected with 2.5 µg of either DBH(232)-Luc or mutant reporters with or without 0.5 µg of RSV-PKA by the calcium phosphate method. HepG2 cells were similarly transfected with 2.5 µg of either DBH(232)-Luc or mutant reporters, with or without 0.25 µg of RSV-PKA, along with 0.25 µg of either control or HA-Arix expression vector as indicated. All transfections included 0.5 µg of pRL-null, which expresses Renilla luciferase, as a control for transfection efficiency. Cells were harvested 48 h after transfection, and lysates were subjected to luciferase assay (Dual-Luciferase Assay System, Promega). Luciferase activity for each reporter is expressed as FL/RL units, meaning reporter FL activity relative to the control RL activity. Each bar represents the mean ± S.E. from 3-6 independent transfections.

To evaluate the importance of the homeodomain sites to the activity of the DBH promoter in the SHSY-5Y neuroblastoma cells, the DBH(232) reporter constructs shown in Fig. 1A were transfected in the presence or the absence of a PKA expression vector. In SHSY-5Y cells, the wild-type DBH(232) reporter exhibited substantial basal activity and a 3-fold induction by PKA (Fig. 1B). Mutations of either the proximal HD3 or the distal HD1/2 pair reduced basal promoter activity to 18 and 6% of wild type activity, respectively. DBH promoter activity stimulated by PKA is also reduced by mutation of homeodomain sites, although the -fold change in promoter activity remains equivalent to the wild type promoter. The combination of the HD3 mutation with the HD1/2 mutations (3HDm reporter) further reduces luciferase expression to 3 and 1% of wild-type levels for basal and PKA treatment, respectively. The severe reduction in activity with each HD site mutation suggests that these sites are interdependent and function to synergistically activate the promoter maximally.

To compare the regulatory components of the DBH promoter between neuronal and non-neuronal cell lines, similar experiments were performed in the liver-derived HepG2 cell line. As previously reported, the basal activity of the DBH promoter is 60-fold stronger in the neuronal cell line than in HepG2 cells (Fig. 1, B and C). In SHSY-5Y cells, the combined HD site mutations reduced the overall reporter activity to levels comparable with those observed with the wild-type promoter in HepG2. These HD mutations had little effect on the low basal or PKA-mediated activity in HepG2 cells. This finding suggests that the HD sites may function in a tissue-specific manner to facilitate strong DBH promoter activity in SHSY-5Y cells.

To evaluate the individual contributions of the HD1/2 and HD3 sites as Arix-responsive regulatory elements, we examined the effects of the HD3 and HD1/2 site mutations on Arix- and PKA-mediated regulation of the DBH promoter in HepG2 cells. Fig. 1C illustrates the previously reported observation (9) that co-transfection of either Arix or PKA expression constructs with the DBH(232)-Luc construct results in a 4-6-fold stimulation of reporter gene activity, while together these proteins synergistically interact to drive activity nearly 50-fold greater than DBH(232)-Luc alone. Although the HD site mutations had little effect on promoter activity in the absence of Arix, these mutations did alter the synergistic activity of Arix both in the presence and the absence of PKA. When the DB1-associated HD pair (HD1/2) is mutated, basal activation by Arix is reduced to 55% of the wild-type promoter. Furthermore, the synergistic activity of Arix and PKA is dropped to 40% of that of DBH(232)-Luc. Mutation of the promoter-proximal HD3 site also resulted in a sizable reduction of Arix-stimulated activity. Disruption of this site alone (HD3m) or in combination with the HD1/2 sites (3HDm) effectively eliminated Arix responsiveness, reducing basal activity by 85% and the synergistic activation with PKA by 96%. These findings indicate that the multiple HD sites in the DBH promoter each is critical for the positive regulation of the DBH gene by Arix. The effectiveness of the HD1/2 sites is dependent upon the presence of an intact HD3 site, suggesting functional interdependence of the homeodomain recognition sites.

Arix Has Distinct Functions at Different HD Sites in the DBH Promoter-- To further distinguish between the individual contributions of HD3 from that of HD1/2 in the regulation of the DBH promoter by Arix, we utilized the Gal-DBD:Gal4-UAS mammalian one-hybrid promoter system. The promoter-distal HD1/2 elements were substituted with a single Gal4-UAS element. This element allows for the recruitment of fusion proteins containing the Gal-DBD to the distal DB1 enhancer in lieu of the HD1/2 elements. By fusing Arix to the Gal-DBD, we can evaluate the contribution of Arix or subdomains of Arix to transcription at the DB1 site. Using various mutant promoters, we can differentiate between activation at the DB1 enhancer through the Gal4-UAS and activation through the natural HD3 site. In addition, we can ensure that Arix, but not other homeodomain protein, is binding to the HD1/2 sites in SHSY-5Y cells.

Constructs fusing the Gal-DBD to Arix were made and tested for their ability to promote transcriptional activation of the DBH promoter in SHSY-5Y cells. In the construct containing an intact HD3, but with HD1/2 substituted with Gal4-UAS (2HDgal), Gal-Arix, but not Gal-DBD alone, increased both basal and PKA-stimulated activity of the DBH promoter (Fig. 2A). In the construct containing both a mutant HD3 and HD1/2 substituted with Gal4-UAS (3HDgal), basal activity is reduced to 25% of the parental vector. However, recruitment of Gal-Arix to the DB1 enhancer increases both basal and PKA-stimulated transcription. When Gal-Arix is added, the relative increase in activity induced by PKA (-fold basal) is doubled in comparison with control vectors. This -fold increase in PKA activity induced by Gal-Arix was observed with either reporter and, thus, is independent of HD3. These results demonstrate that in SHSY-5Y cells, recruitment of Arix to the DB1 regulatory element can potentiate PKA stimulation, even when the promoter-proximal homeodomain binding site is missing. However, the promoter-proximal HD3 is necessary for maximal transcriptional activity.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Recruitment of Gal-Arix to mutant HD1/2 elements restores Arix basal and PKA-mediated DBH transcription in both SHSY-5Y and HepG2 cells. The 2HDm and 3HDm mutant promoters were modified by replacing the DB1-related HD sites with a single Gal4-UAS (2HDgal and 3HDgal). The promoter schematic at the left illustrates the HD site mutations (striped boxes) and Gal-UAS substitutions (large box) in the respective promoters. The Gal-UAS serves to recruit Gal-DBD fusion proteins to the promoter. SHSY-5Y (A) and HepG2 cells (B) were transiently transfected with 2.5 µg of mutant reporter construct, 0.5 µg of pRL-null, with or without 0.25 µg of RSV-PKA along with 0.5 µg of the control expression vector, wild-type HA-Arix (HepG2 cells only), the control Gal-DBD construct, or the Gal-Arix construct as indicated. Luciferase activity for each transfection was assayed and expressed as described in the legend to Fig. 1. The relative PKA stimulation for each activator/reporter combination represents the -fold increase in PKA-stimulated activity expressed relative to its basal activity. Each bar represents the mean ± S.E. from 6-9 independent transfections.

Experiments were then performed in the HepG2 cell line, where either Arix or Gal-Arix can be individually added to the system. When the HD3 site was present in the mutant promoters (i.e. 2HDm and 2HDgal), both Arix and Gal-Arix stimulated basal activity equally well, exhibiting a 4-6-fold increase compared with control vectors (Fig. 2B). Wild-type Arix activity was eliminated by the additional mutation of the HD3 site in the 3HDgal and 3HDm reporters. Gal-Arix, however, retained approximately 50% of its ability to activate the DBH promoter in the 3HDgal construct. These results indicate that recruitment of Arix to the DB1 enhancer through the Gal-UAS can provide moderate basal activation. However, the HD3 is necessary for the majority of the basal activity elicited by Arix or Gal-Arix.

The influence of the different Arix binding sites on PKA stimulation of transcription was also evaluated. Insertion of the Gal-UAS in the DB1 homeodomain sites reduced PKA-mediated activity by wild-type Arix as compared with Gal-Arix. This result indicates that PKA-stimulated activity from either the 2HDgal or 3HDgal reporters was maximal only when Gal-Arix is recruited to the DB1 enhancer and not when Arix is recruited to the HD3 site. In theory, the 2HDgal promoter should function similarly to the wild-type DBH(232) promoter, recruiting Arix to both the DB1 enhancer and the HD3 site. Indeed, the activation of the 2HDgal reporter by Gal-Arix did recapitulate the activation levels of Arix on the wild-type DBH(232) promoter (compare Fig. 1C with Fig. 2B).

These analyses indicate that Arix acts as a transcriptional activator through multiple HD sites in the DBH promoter. The results of these experiments suggest that the promoter-proximal HD3 is critical for basal activation of the DBH promoter by Arix, while the promoter-distal HD1/2 functions to a greater extent to influence PKA-stimulated promoter activity. Since HD1/2 is adjacent to the functional PKA-responsive CRE/AP1 site, it is possible that the binding of Arix to HD1/2 physically influences recruitment of the AP1 proteins to the CRE/AP1 site. In the promoter-proximal position, HD3, Arix may interact with the general transcriptional machinery to stimulate DBH transcription. The distinct contexts provided by the different Arix binding sites appear to dictate the mechanism by which Arix drives DBH promoter activity.

Arix-Arix Interactions May Influence Transcription-- Fig. 2 demonstrates that recruitment of Arix to DB1 homeodomain sites (HD1/2) potentiates PKA-stimulated transcription, even in the absence of the HD3 binding site. However, results in Fig. 1 indicate that the loss of HD3 reduces the response to Arix and PKA. These experiments are apparently inconsistent in that one experiment suggests a dependence on HD3 for all DBH transcription, while the other suggests an independent function for HD1/2 in the absence of HD3. One hypothesis to explain this discrepancy is that binding of Arix to homeodomain sites may be influenced by Arix-Arix protein interactions, while the binding of Gal-Arix to Gal-UAS is not.

To investigate whether Arix interacts with itself, we used the mammalian two-hybrid system. The reporter used in these analyses, 5Gal-DBH(62)-Luc (Fig. 3), places five copies of the Gal4 DNA recognition site (Gal-UAS) directly adjacent to a minimal DBH promoter (62/+10) to drive expression of luciferase. This minimal DBH-(62/+10) promoter does not contain the proximal HD3 site but places the 5×Gal cassette at approximately the same position as the HD3 in the DBH(125)-Luc construct. Co-transfection of this reporter construct with Gal-Arix into SHSY-5Y cells results in a modest stimulation of reporter gene activity, while transfection of the construct containing the Gal-DBD alone does not influence activity (Fig. 3). A second Arix expression construct, Arix-VP16, consists of Arix fused to the transcriptional activation domain of Herpes simplex virus protein, VP16. When Arix-VP16 is co-transfected with 5Gal-DBH(62)-Luc, there is no stimulation of transcription, because there is no Arix DNA-binding domain in the reporter construct. However, when Arix-VP16 plus Gal-Arix is co-transfected with 5Gal-DBH(62)-Luc, transcription is stimulated 6-fold over that with Gal alone. These results suggest that Arix-Arix interactions occur within the neuroblastoma cell, leading to the recruitment of the VP16 activation domain to the DBH promoter. These Arix-Arix interactions could potentially influence binding of Arix to the multiple homeodomain binding sites on the genomic DBH promoter.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Arix-Arix interactions may influence transcription. The 5Gal-DBH(62)-Luc reporter construct consists of a 5×Gal-UAS cassette placed directly adjacent to a minimal DBH promoter 62 nucleotides upstream of the transcription start site. 1 µg of reporter plasmid was co-transfected into SHSY-5Y cultures along with 1 µg of the Gal-DBD fusion constructs and 1 µg of the VP16 fusion construct, as indicated. Under these conditions, transcription from the promoter is stimulated only when interaction occurs between the protein bound to the Gal4 sites and the protein containing the VP16 activation domain. Values reported represent the mean ± S.E. of triplicate samples from a single experiment that has been repeated three times.

The N Terminus of Arix Contains an Activation Domain-- The results presented in Fig. 1 indicate that the HD3 site provides a functional element for Arix-mediated transcriptional activation. To define the activation domain(s) of Arix when bound to this single functional HD site, we used a truncated DBH promoter containing the HD3 (DBH(125)-Luc) (Fig. 4A). A series of Arix expression constructs were made in which a portion of the Arix protein was deleted in relation to the more centrally located DNA-binding homeodomain (Fig. 4). Each of these constructs contains the intact DNA-binding domain of Arix and thus should bind to the HD3 in the DBH(125) promoter construct.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Arix stimulates the proximal DBH promoter through an N-terminal activation domain. A, a schematic representation of the two promoters used to define the minimal activation domain necessary to drive basal transcription from the DBH proximal promoter. The DBH(125)-Luc reporter construct consists of a minimal DBH promoter containing 125 nucleotides upstream of the transcription start site and includes the HD3 site (white box). The 5Gal-DBH(62)-Luc reporter construct consists of a 5×Gal-UAS cassette placed directly adjacent to a minimal DBH promoter 62 nucleotides upstream of the transcription start site. B, the DBH(125)-Luc reporter was used to test the transcriptional activity of truncation constructs of Arix. HepG2 cells were transfected with 0.5 µg of DBH(125)-Luc and 0.5 µg of pRL-null along with control vector, full-length HA-Arix, or HA-tagged Arix truncations as indicated. C, the 5×DBH(62)-Luc reporter was used to further define the activation domain of Arix. Gal-DBD fusion constructs with full-length Arix or Arix truncations were used to recruit the putative activation domain to the Gal-UAS sites in the heterologous promoter. Luciferase activity for each transfection was assayed and expressed as described in Fig. 1. Each bar represents the mean ± S.E. from 3-6 independent transfections.

When transfected into HepG2 cells, Arix strongly activates the DBH(125) promoter, producing activity nearly 10-fold greater than the control expression vector (Fig. 4B). The ArC (aa 1-151) construct, deleted of all residues C-terminal to the homeodomain, elicited even greater activity, more than 2-fold greater than wild-type Arix and 25-fold greater than control. Deletion of amino acids N-terminal to the homeodomain, ArN (aa 84-281), reduced activity by 50% compared with full-length Arix. Activity elicited by the construct containing only the homeodomain and surrounding amino acids, ArHD (aa 84-151), was severely compromised compared with wild-type Arix and was roughly equivalent to the control vector. These results indicate that the homeodomain is not sufficient for activation from the HD3 site. The N-terminal portion of Arix, however, appears necessary for maximal activation of the proximal DBH promoter. Therefore, the N-terminal polypeptide probably contains all or part of an activation domain. Furthermore, the C-terminal domain appears to negatively regulate the function of the activation domain. The results of this Arix truncation analysis on the proximal DBH(125) promoter are similar to those previously seen using the longer DBH(232) promoter,² confirming the requirement of the Arix N terminus for DBH promoter activation.

These results demonstrate that the N terminus is necessary for basal activation, yet they do not address whether this portion of the Arix protein is sufficient for basal activation in the absence of the DNA-binding homeodomain. In order to further define the putative activation domain of Arix, we used the heterologous promoter Gal4-UAS system to recruit portions of Arix to the promoter, independent of the DNA binding function of its homeodomain. The reporter used in these analyses, 5Gal-DBH(62)-Luc, places a 5×GAL4-UAS cassette directly adjacent to a minimal DBH promoter (62/+10), to drive expression of luciferase.

Expression constructs containing in-frame Gal4-DBD fusions with full-length Arix or portions of Arix were transfected into HepG2 cells along with the 5Gal-DBH(62)-Luc reporter. Wild-type Arix is only as effective as control vectors (pcDNA3 or Gal-DBD vector) at activating the 5Gal-DBH(62)-Luc reporter, which lacks all three HD binding sites (Fig. 4C). Fusing the Gal4-DBD to full-length Arix, so that Arix is recruited to the promoter, resulted in a 7-fold transactivation of the promoter. As with the DBH(125) promoter, deletion of the C-terminal amino acids of the Gal-Arix fusion (Gal-ArC, Arix aa 1-151) resulted in a substantial increase in activation compared with full-length Gal-Arix, confirming the regulatory nature of the C-terminal domain. Most notable, however, is that recruitment of the N-terminal domain alone (Gal-NAr, Arix aa 1-100) to the 5Gal-DBH(62) promoter activated transcription to an extent nearly twice that of Gal-Arix activity. Thus, the N-terminal 100 amino acids alone are sufficient to activate transcription, even in the absence of the homeodomain. Neither the homeodomain alone (Gal-ArHD, Arix aa 84-151) nor the C-terminal Arix (aa 151-281) domain (Gal-CAr) exhibits transcriptional activity when recruited to the 5Gal promoter. These results identify an activation domain in the N-terminal 100 amino acids of Arix that is necessary and sufficient for activation of the basal DBH promoter.

The Arix N Terminus Has a Distinct Function at the DB1 Enhancer-- The results of the preceding experiments suggest that the location of the HD3 site serves to recruit the N-terminal activation domain of Arix to a promoter-proximal position. In this location, approximately 30 bp upstream of the TATA box, it appears that the function of this domain is to regulate the overall gain of promoter activation. The more distally located DB1-related HD sites are also critical for maintenance of basal DBH promoter activity and additionally are essential for promoter modulation by PKA. We therefore tested whether the same domain of Arix, which is necessary for mediating basal transcriptional activation through the HD3 site, is also functional when recruited to the DB1 site.

To perform this analysis, the 3HDgal construct, containing the HD3 mutation and the Gal-UAS substituted for the HD1/2 sites, was used in order to isolate the influence of recruiting Arix to the DB1 enhancer region (Fig. 5). Transfection of the 3HDgal reporter construct with either Gal-Arix and Gal-ArC stimulated basal promoter activity. In contrast to the strong activation on the 5Gal-DBH(62)-Luc by Gal-NAr, recruitment of this N-terminal domain to the distal DB1 site was not sufficient to drive basal activation of 3HDgal. These findings suggest that basal activation through the DB1 enhancer site requires both the N terminus and the intact homeodomain.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   The basal and PKA modulatory functions of Arix through the DB1 enhancer require the N terminus. The 3HDgal reporter construct was used to define the Arix domains necessary for DB1 enhancer dependent regulation of basal and PKA-mediated transcription. The Gal-UAS site recruits the Gal-DBD fusion proteins to the DB1 enhancer region. HepG2 cells were transfected with 0.5 µg of the 3HDgal reporter with or without 0.25 µg of RSV-PKA and 0.5 µg of pRL-null along with 0.25 µg of either control vector or Gal-DBD fusion constructs as indicated. Luciferase activity for each transfection was assayed and expressed as described in the legend to Fig. 1. Each bar represents the mean ± S.E. from three independent transfections.

The ability of the Gal fusion proteins to stimulate transcription in the presence of PKA was also evaluated. In these experiments, all constructs containing the N-terminal 100 amino acids of Arix were able to stimulate PKA responsiveness of the DBH promoter. Thus, the minimal domain of Arix necessary for maximal PKA activation from the DB1 site lies between amino acids 1 and 100. The finding that the Gal-NAr construct does not stimulate basal activation from the DB1 site, but is sufficient for PKA activation, suggests that there may be multiple functional domains within this 100-amino acid segment of Arix. The results of these analyses are summarized in Fig. 6A.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   A, a schematic summary of the minimal functional domains necessary to drive transcription from the proximal HD3 and distal DB1 enhancer regions of the DBH promoter. B, an alignment of homologous segments of a Brachyury-like motif contained within the N-terminal activation domain of Arix (residues 61-75). This motif was found in the Arix family member, NBPhox/Phox2b, the T-box transcription factor, Brachyury, and the paired-box transcription factor, Pax9. The Brachyury-like motif is conserved in transcription factors from several divergent classes and may have relevance in activation domain function. Presented for alignment are the sequences for rat Arix, mouse Phox2b, pig and mouse Brachyury, and mouse Pax9.

Since the N-terminal domain of Arix is necessary for the transcriptional activity of Arix, a BLAST search was performed to determine whether the N-terminal activation domain of Arix was homologous with other putative transcriptional activation domains. Using the NCBI online PSI-BLAST search engine (available on the World Wide Web), we searched the GenBank^TM data base for sequence homologies using amino acids 1-89 of rat Arix. As expected, this search identified the N terminus of NBPhox as being 56% identical and 70% conserved compared with the N terminus of Arix. In addition, this search identified segments in the T-box protein, Brachyury, and the paired box protein, Pax9, that had high homology with Arix (Fig. 6B). A 14-amino acid stretch of Brachyury and a 13-amino acid stretch of Pax9 aligns with a segment of Arix, which is located near the homeodomain (aa 64-73). The finding of this conserved sequence in transcription factors of different classes may be indicative of a common functional domain or motif. Further analysis is needed to determine if this subdomain is critical for transcriptional activity of Arix.

The Coactivator CBP Functionally Augments Arix-mediated Transcription and Physically Interacts with the N Terminus of Arix-- CBP/p300 family coactivator proteins play a critical role in activator-dependent transcription through their interaction with transcription factors, including members of the CREB and AP1 families. The findings that Arix facilitates transcription through multiple HD sites in the DBH promoter and interacts with AP1 proteins through the CRE/AP1 site of the DB1 enhancer led to the hypothesis that CBP may integrate the multiple transcription factor binding sites by functioning as a coactivator in Arix-regulated DBH transcription. We explored this hypothesis by measuring the effects of Arix and CBP on regulated expression of the DBH(232)-CAT reporter in HepG2 cells. In the absence of Arix, transfection of the RSV-CBP expression construct had no effect on basal activation of the DBH promoter in HepG2 cells and elicits a modest 2-fold increase in PKA-stimulated activity (Fig. 7). Arix exhibited a dose-dependent increase in basal activation of the DBH-CAT promoter, exhibiting a 3.5-fold increase at the highest amount tested. CBP had a subtle effect on the basal activity mediated by Arix, inducing a 1.5-2-fold increase in the promoter activity at the highest Arix doses. However, effects of CBP cotransfection are most dramatic in the presence of Arix plus PKA, where CBP synergistically augmented the response of the DBH promoter to Arix plus PKA. Even at the lower amounts of Arix, where the presence of CBP elicited little or no effect on basal DBH promoter activity, substantial synergistic activation occurred in the presence of CBP, Arix, and PKA. For example, at 0.2 µg of Arix, CBP has no effect on basal promoter activity but raises the PKA-induced activity from 75- to 400-fold greater than basal. This dose-response relationship suggests that the amount of CBP present in the cell is limiting the Arix plus PKA responsiveness of the DBH promoter, and that CBP may interact in conjunction with Arix and the PKA-responsive AP1 proteins at the DB1 enhancer to stimulate transcription of the DBH promoter.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   The coactivator CBP augments Arix regulation of the DBH promoter in a PKA-dependent manner. HepG2 cells were cultured in 100-mm plates and transfected by the calcium phosphate method with 5 µg of DBH(232)-CAT reporter, 2 µg of RSV-Luc, ± 5 µg of RSV-CBP, with or without 2.5 µg of RSV-PKA, and varying amounts of RSV-Arix expression vector. The total amount of expression vector was equalized to 20 µg with the backbone expression vector. Cells were harvested 48 h after transfection, and lysates were assayed for CAT and luciferase activities. Activity values were calculated relative to the total amount of protein in the lysate and normalized relative to the basal promoter activity in the absence of PKA, CBP, and Arix. Thus, the mean value of -fold CAT activity for basal control is equal to 1. Values represent mean + S.E. from six independent transfections.

We next wanted to determine whether the functional stimulation of Arix by CBP resulted from a direct physical interaction between these two proteins. The CBP protein possesses several modular domains that are known to interact with various transcription factors. Several of those domains, produced as GST fusion proteins, were tested for their ability to interact with Arix. Each CBP domain fusion protein was incubated with radioactively labeled Arix in an in vitro interaction assay, as illustrated in Fig. 8A. Arix exhibited a strong specific interaction with only CBP (1679-1874). This domain contains the third zinc-finger motif (C/H3).

View larger version (44K): [in this window] [in a new window]   Fig. 8.   The N-terminal domain of Arix is necessary for a direct physical interaction with the C/H3 domain of CBP. A, a schematic diagram of the CBP protein. Specific structural domains are indicated in shading, including the CREB binding domain (CBD), putative zinc finger domains (C/H1, C/H2, and C/H3), and the bromo domain (Br). Boxes below designate the CREB binding domains used as GST fusion proteins for the present analyses. B, in vitro translated (IVT) 35S-labeled Arix was incubated with glutathione-agarose beads containing GST fusion proteins of CBP domains or GST alone as indicated. After extensive washing and SDS-polyacrylamide gel electrophoresis, bound Arix was detected by autoradiography. C, Arix truncations (Arix-(1-151) and Arix-(84-281)), full-length Arix-(1-281), and full-length NBPhox-(1-314) were in vitro translated and incubated with GST-CBP-(1679-1874). After incubation, beads were washed and prepared for SDS-polyacrylamide gel electrophoresis analysis and autoradiography.

Further analysis was carried out to determine which portion of Arix was necessary for the interaction with this C/H3 domain. The N-terminal and C-terminal truncation constructs, identical to those used in Fig. 4 (ArC contained aa 1-151, and ArN contained aa 84-281) were in vitro translated and incubated with either GST-CBP (aa 1679-1874) or GST alone. Full-length Arix and Arix (aa 1-151), bound strongly to the C/H3 domain of CBP, while the Arix (aa 84-281) construct did not bind to this domain. This suggests that the N-terminal arm of Arix is necessary for the interaction with CBP in vitro. Additionally, we tested whether NBPhox, another member of the Arix/Phox2 homeodomain protein family, interacted with CBP. NBPhox also interacted specifically with the C/H3 domain. The homeodomains of Arix and NBPhox are identical, and the N-terminal domains are highly homologous (50% shared amino acids) as well. The present findings indicate that the N-terminal activation domain may exert some of its activational ability by directly interacting with CBP to stably recruit it to the DBH promoter.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The differentiation and maintenance of the noradrenergic phenotype is guided by the interaction of exogenous cues with intrinsic genetic machinery to activate the expression of the catecholaminergic biosynthetic enzymes (3). The present study supports the hypothesis that the paired-like homeodomain protein, Arix, acts as a critical phenotype-specific regulator of the DBH promoter by serving as an integrator of signal-dependent transcription activators within the network of the general transcription machinery.

In the present study, we demonstrate that Arix directly activates the DBH promoter through multiple HDs located in the DB1 enhancer and proximal promoter. In the catecholaminergic neuroblastoma cells, the distal and proximal HD sites are each essential for transcriptional activity of the DBH promoter. Recruitment of Arix to the distal HD pair, adjacent to the CRE/AP1, is necessary for maximal response to PKA, while Arix activation through the proximal HD3 element enables general promoter activation. The coactivator, CBP, appears to be a functional component of the signal-dependent transcriptional network involved in Arix-mediated PKA activation. Taken together, these findings suggest a potential model for Arix dependent activation of the DBH gene, in which Arix serves as the necessary functional link between the basal transcription apparatus and signal-dependent transcription machinery (outlined in Fig. 8). Since recent reports indicate that the closely related transcription factor NBPhox/Phox2b appears to be functionally interchangeable with Arix/Phox2a (11),² it is likely that the functional characteristics of Arix outlined in the present study are also shared by NBPhox. The finding that NBPhox physically interacts with the same domain of CBP as Arix supports this notion.

Arix Functions as a Tissue-specific Regulator through Multiple HD Sites-- Previous studies have shown that the HD elements within the DBH promoter can serve as both binding sites and transcriptional regulatory elements for the members of the Arix/Phox2 homeodomain family (8-12). Our present results extend the previous findings by demonstrating that not only are the proximal HD3 and distal HD1/2 sites necessary for basal responsiveness of the DBH promoter in a DBH-positive cell line, but each element is also are critical for the full expression of the PKA-dependent promoter stimulation. In addition, our studies demonstrate that within the context of the DBH promoter, these sites function as Arix-responsive elements. Not only are these sites targets of Arix in noncatecholaminergic cell lines, such as HepG2, but direct recruitment of Arix to the HD1/2 sites in the neuroblastoma cell line, via the Gal-UAS, restores transcriptional activity from the DBH promoter. While it might be predicted that Arix function is dependent upon binding to specific recognition sites, recent studies on the related Phox1 (33) and NBPhox (34) transcription factors suggest that DNA binding may not be required for transcription factor function. Our results demonstrate that the function of Arix as a tissue-specific activator of DBH expression is dependent on its recruitment to multiple sites within the DBH promoter.

Arix Has Distinct Functions through Different HD Sites-- The functional roles of Arix at different HD elements appear to be interdependent in that neither site in the absence of the other conveys the full basal or PKA responsive function of the wild-type promoter (Fig. 1, B and C). One possible explanation for the interdependence of sites may lie in protein-protein interactions. The Arix-Arix interactions, demonstrated in Fig. 3, suggest that Arix bound at one site may potentiate binding at the other sites. Elimination of one binding site by mutation may influence overall binding at the other sites, thus diminishing transcriptional activation by Arix.

The promoter-proximal position of the HD3 site suggests that it may function to facilitate interaction with the general transcriptional apparatus, while the location of the HD1/2 sites, adjacent to the CRE/AP1 site, suggests involvement in the response to PKA. When a pair of HD sites in the DB1 is mutated, the -fold induction of reporter gene activity by PKA is diminished by 2-fold. The full induction is restored when Arix is recruited to the promoter, using the Gal-Arix constructs, suggesting that binding of Arix at HD1/2 plays a predominant role in the stimulation of DBH promoter activity by PKA.

The CRE/AP1 site, located adjacent to the HD1/2 element, is essential for PKA-induced activation as well as the recruitment of Fos/Jun family members to the DBH promoter (10). The CRE/AP1 site is also necessary for the synergistic activation of the DBH promoter by Arix and PKA (9). The close proximity of these two elements in the DB1 enhancer and the functional synergism of the proteins binding to each site suggest that Arix may cooperatively interact with members of the AP1 family. However, experiments designed to detect the physical interaction between Arix and c-Fos or c-Jun have been negative.³ Alternatively, the binding of Arix to the DB1 enhancer may provide a favorable environment for the recruitment of Fos-Jun heterodimers to the CRE/AP1 site following PKA stimulation. This interaction could take the form of a conformational change in the AP1 recognition site resulting from Arix binding to the adjacent HD elements. A comparable role for the paired-like homeodomain protein Phox1 has been suggested, where Phox1 enhances the binding of the serum response factor to the serum response element of the c-fos gene (35). The cooperative interaction of Phox1 and serum response factor was also reported to facilitate the recruitment of the extracellular signal-responsive serum response factor accessory protein, Elk1, subsequently enhancing transcription from the c-fos serum response element (35). The interactions described imply a physical contact between these proteins, yet in vitro assays do not demonstrate stable complex formation of Phox1 with either serum response factor or Elk1. Thus, the binding of Arix and Phox1 homeodomain proteins to the DNA may enhance transcriptional function of other factors by altering DNA structure in such a manner as to promote DNA-protein interaction.

The N-terminal Portion of Arix Functions as an Activation Domain-- To address the molecular mechanisms underlying Arix regulation of the DBH gene, a functional domain analysis of Arix was performed. The hybrid DBH promoters incorporating the Gal4-UAS recruitment of GAL-DBD fusion proteins provided a system to identify the domains critical for Arix-mediated activation within the context of the DBH promoter. The hybrid system allowed us to differentiate critical domains acting through the proximal HD3 site and the distal DB1 enhancer. As summarized in Fig. 6, our results indicate that the N-terminal portion of Arix (aa 1-100) contains an activation domain that is necessary to drive basal transcription when recruited to the proximal DBH promoter. When this same N-terminal domain was recruited to the distal DB1 enhancer location, it was sufficient to activate a PKA response from the promoter but did not elicit basal activation. The intact homeodomain was required along with the N terminus to elicit basal activation through the DB1 enhancer. These findings suggest that the basal activation by Arix through the HD3 site involves molecular interactions with the general transcriptional apparatus distinct from those required to elicit the more modest basal activation from the DB1 enhancer. Furthermore, the molecular interactions regulating PKA-mediated activation through the DB1 enhancer are sufficiently carried out by the N-terminal portion of Arix, strongly implicating the involvement of CBP interaction with this domain of Arix (as discussed below).

Finding that the deletion of the C-terminal section of Arix increases transcription factor activity suggests that this C-terminal region may negatively regulate Arix activity in the intact protein. There are several consensus phosphorylation sites in this domain, and it can be phosphorylated in vitro.⁴ Perhaps one component of the Arix plus PKA stimulation of DBH transcription lies in modification of the C-terminal domain to relieve repression and enhance Arix function. Alternatively, protein-protein interaction between the C-terminal domain and other co-regulatory proteins may influence activity.

A BLAST search using N-terminal amino acids (aa 1-89) of Arix revealed that a 15-amino acid stretch (aa 61-75) exhibited high homology with a member of the T-box gene family, Brachyury, and the paired-box protein, Pax9. This short stretch of homology was given the name the Brachyury-like motif to reflect the abundance of homologues with this motif. Both Brachyury and Pax9 are putative transcription factors that play important roles in early development. Brachyury is required for the formation of posterior mesoderm and for the axial development (36). Pax9 plays an essential role during the development of organs derived from endoderm, mesoderm, and neural crest (37). Functional domain studies of the Brachyury protein of mouse and Xenopus and Pax9 from zebrafish indicate that the Brachyury-like motif is located between the transactivation and the DNA-binding domains in each case (38-40). Although this motif within Brachyury or Pax9 does not appear to be directly involved in either transcriptional activation or DNA binding, it is conserved among a variety of different species, suggesting that this short peptide stretch may be a structural motif critical in orienting the transactivation and the DNA-binding domains of transcription factors. This organization is also found in Arix and NBPhox, where the N-terminal activation domain and the DNA-binding homeodomain are bridged by the Brachyury-like motif. These observations suggest conservation of this structural organization across transcription factors of different classes.

The PKA-dependent Interaction of Arix and CBP Suggests Cooperative Recruitment to the Promoter-- The interaction of Arix/Phox2 proteins and the cAMP/PKA pathway is critical for the developmental regulation of tyrosine hydroxylase and DBH expression in the neural crest progenitor cells (17). Our investigations have focused on determining the mechanisms that underlie the synergistic interaction of Arix with the PKA pathway. Our previous studies have shown that the CRE/AP1 site is critical for the response to PKA, and that proteins bound to that site change from a Jun family complex to Fos-Jun heterodimers upon PKA activation (10). Now we demonstrate that Arix contacts the transcriptional coactivator, CBP, and suggest that CBP forms a bridge between proteins bound at the DB1 enhancer and the general transcriptional machinery. CBP has also been implicated in the cAMP-mediated activation of the tyrosine hydroxylase gene, encoding the protein responsible for the biosynthesis of dopamine (4). These studies suggest that CBP integrates the signal transduction pathway with the tissue-specific homeodomain proteins to induce developmentally programmed gene expression. The observation that the addition of CBP to cells stimulates gene transcription suggests that intracellular CBP is limiting and that there is competition among many active genes for available CBP.

The transcriptional coactivators, CBP/p300, have been extensively studied in relation to signal-dependent transcriptional activation. They have been shown to physically interact with a number of transcriptional activators that have been demonstrated to bind to the DBH promoter including CREB, (20), c-Fos (21), and c-Jun (22). Our results demonstrate that Arix also makes a contact with CBP through the interaction between the N-terminal activation domain of Arix and the third zinc finger domain of CBP, which is the same domain that binds to c-Fos (21). c-Jun interacts with an N-terminal domain of CBP, within the same region that associates with CREB (22). The modular nature of interaction domains on CBP suggests a model by which multiple activators, distributed along the DBH promoter, may form multiple contacts with CBP to facilitate PKA-dependent transcription (Fig. 9). These multiple contacts serve to stabilize the recruitment of CBP to the promoter, where it can then facilitate transcription through chromatin remodeling and its interaction with the basal transcription apparatus.

View larger version (32K): [in this window] [in a new window]   Fig. 9.   A summary model of the prospective regulatory machinery involved in basal and PKA-regulated expression of the DBH promoter. The evidence presented here suggests that the interaction of Arix at multiple sites within the DBH promoter is necessary to drive both basal and PKA-stimulated transcription from the promoter. In the absence of Arix, the DBH promoter is minimally activated, suggesting that interaction of basal transcriptional machinery with the promoter-proximal factors requires the presence of Arix. When Arix is present, the promoter is competent to drive transcription in conjunction with other essential promoter elements, such as Sp1. PKA stimulation promotes the change of AP1 transcription factors binding to the CRE/AP1 site and together with Arix provides multiple contact sites for recruitment of the coactivator CBP. The requirement of the CRE/AP1 as well as the distal and proximal HD sites suggests that PKA-induced transcriptional activity results from the integration of signal-dependent transcription machinery with the basal transcription apparatus.

The transcriptional studies demonstrate that CBP plays a role in the stimulation of DBH transcription primarily when both Arix and PKA are present. There are several possible mechanisms by which the interaction between CBP, Arix, c-Fos, and c-Jun could be dependent upon PKA. First, cAMP induces the expression of c-Fos in catecholaminergic cells (10), allowing the composition of factors bound to the AP1 site to change from the Jun family to Fos-Jun heterodimers. It is possible that the Fos-Jun-Arix complex on the DB1 enhancer engages in more stable contacts with CBP than the complex present under nonstimulated conditions. These multiple contacts may lead to stabilization of CBP binding, recruiting the limiting CBP away from other genes to the DBH gene, as has been suggested in other models of CBP function (42). A second possible mechanism involving the PKA dependence of CBP action may be the need for phosphorylation of c-Jun, c-Fos, or Arix. Initially, it was reported that phosphorylation of c-Jun at Ser⁶³ and Ser⁷³ by the N-terminal Jun kinase was essential for both Jun-CBP contact and subsequent stimulation of gene expression (43). Subsequent studies have found that phosphorylation of neither c-Jun nor c-Fos is required for activation by CBP (44). Arix can be phosphorylated in vitro,⁴ but the functional consequences of phosphorylation are unknown. The possible relationship between Arix phosphorylation and CBP interaction is open for further investigation. A third possible mechanism to explain the interaction between PKA and CBP involves direct activation of CBP by PKA. In the catecholaminergic PC12 cell line, the N-terminal transcriptional activation domain of CBP is activated by treatment with PKA (45), while in a different neuroendocrine cell line, AtT20, the C-terminal glutamine-rich domain is activated by cAMP (46). CBP may be phosphorylated by kinases activated through the PKA pathway, leading to the stimulation of cAMP-dependent gene transcription in a cell type-specific pattern.

The tissue-specific transcription factors, Arix and NBPhox, are essential for development of the catecholaminergic neuron (4-6). The results of our experiments indicate that interaction of Arix with multiple sites at the DBH promoter serves the biological function of making the neuroblast competent to respond to environmental factors by integrating signal-dependent transcription factors with the general transcriptional machinery.

	ACKNOWLEDGEMENTS

We acknowledge the technical assistance of Barbara Mason in portions of this study. We thank Drs. Richard Maurer, Richard Goodman, Paul Shapiro, and Hua Lu for providing us with plasmids used in the experiments described here.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM38696 (to E. J. L.) and a fellowship from the American Heart Association, Oregon Affiliate (to D. J. S.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. Biochemistry and Molecular Biology, Oregon Health Sciences University, L224, Portland, OR 97201. Tel.: 503-494-5076; Fax: 503-494-8393; E-mail: lewis@ohsu.edu.

² M. Adachi, D. J. Swanson, and E. J. Lewis, submitted for publication.

³ D. J. Swanson, M. Adachi, and E. J. Lewis, unpublished observations.

⁴ M. Adachi and E. Lewis, unpublished observation.

	ABBREVIATIONS

The abbreviations used are: DBH, dopamine -hydroxylase; HD, homeodomain core recognition site; CAT, chloramphenicol acetyltransferase; FL, firefly luciferase; RL, Renilla luciferase; DBD, DNA-binding domain; CRE, cAMP-response element; CREB, CRE-binding protein; CBP, CREB-binding protein; PKA, the catalytic subunit of cAMP-dependent protein kinase A; GST, glutathione S-transferase; HA, hemagglutinin; aa, amino acid(s); RSV, Rous sarcoma virus; bp, base pair(s); UAS, upstream activation sequence.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. Kvetnansky, E. L. Sabban, and M. Palkovits Catecholaminergic Systems in Stress: Structural and Molecular Genetic Approaches Physiol Rev, April 1, 2009; 89(2): 535 - 606. [Abstract] [Full Text] [PDF]

				R. Benfante, A. Flora, S. Di Lascio, F. Cargnin, R. Longhi, S. Colombo, F. Clementi, and D. Fornasari Transcription Factor PHOX2A Regulates the Human {alpha}3 Nicotinic Receptor Subunit Gene Promoter J. Biol. Chem., May 4, 2007; 282(18): 13290 - 13302. [Abstract] [Full Text] [PDF]

				N. Antcheva, M. Boniotto, I. Zelezetsky, S. Pacor, M. V. V. Falzacappa, S. Crovella, and A. Tossi Effects of Positively Selected Sequence Variations in Human and Macaca fascicularis {beta}-Defensins 2 on Antimicrobial Activity Antimicrob. Agents Chemother., February 1, 2004; 48(2): 685 - 688. [Abstract] [Full Text] [PDF]

				Y Jiang, T Matsuo, H Fujiwara, S Hasebe, H Ohtsuki, and T Yasuda ARIX gene polymorphisms in patients with congenital superior oblique muscle palsy Br. J. Ophthalmol., February 1, 2004; 88(2): 263 - 267. [Abstract] [Full Text] [PDF]

				J. L. Rychlik, V. Gerbasi, and E. J. Lewis The Interaction between dHAND and Arix at the Dopamine {beta}-Hydroxylase Promoter Region Is Independent of Direct dHAND Binding to DNA J. Biol. Chem., December 5, 2003; 278(49): 49652 - 49660. [Abstract] [Full Text] [PDF]

				S. Dziennis and B. A. Habecker Cytokine Suppression of Dopamine-beta -hydroxylase by Extracellular Signal-regulated Kinase-dependent and -independent Pathways J. Biol. Chem., April 25, 2003; 278(18): 15897 - 15904. [Abstract] [Full Text] [PDF]

				F. Muller and H. Rohrer Molecular control of ciliary neuron development: BMPs and downstream transcriptional control in the parasympathetic lineage Development, March 14, 2003; 129(24): 5707 - 5717. [Abstract] [Full Text] [PDF]

				M. Adachi and E. J. Lewis The Paired-like Homeodomain Protein, Arix, Mediates Protein Kinase A-stimulated Dopamine beta -Hydroxylase Gene Transcription through Its Phosphorylation Status J. Biol. Chem., June 14, 2002; 277(25): 22915 - 22924. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Swanson, D. J. Articles by Lewis, E. J. Search for Related Content PubMed PubMed Citation Articles by Swanson, D. J. Articles by Lewis, E. J. Social Bookmarking                      What's this?